High system load > 100%

I have also high load since I updated to V5.0 :rage:
Keeps rising until Homey is really unresponsive.

Hi, Im brand new to Homey, below 14 days. However I have been pusseled by the high CPU load, as you are. Rebooting helps for some time, however easily goes to beyond 80% again. From time to time killing the updates of my Hue and Tado.
Have tried to delete some app, and it helps a little, however not a lot. And, with no fancy app, then why have a homey at all:-)

However, I have tried to “kill” my iPad/iPhone app evert time I have used it, and seen a effect on CPU. Goes from 80% to below 50% for the average 15 min number.

Maybe not related at all, however the IOS app is running all the time for geo location to work, and it might poll a lot of other information all the time, generating CPU load…

Anybody else with similar expierence?

OK, mine is now solved. It was not the iPad app, however the Hue app. My Hue app updated automatically during today, and the problem was solved right after. My version now is 5.1.1. My previous version was 5.0.4. Along the period I also got 2 updates of Homes, both 5.0.1 and 5.0.2. I beleive the problem is related to versions of Homey vs Hue app

Reducing CPU load % from 150% to 30% on a HP2023.

Although this thread is somewhat older, I would like to share my experience. It will help mainly those with many (motion) devices or time repeating tasks.

In the past weeks I had CPU often at 120-180% leading to longer response times for example for motion controlled lights. I had 25+ Hue motion sensors generating recurring and even almost simultaneous motion events when moving between rooms.

Here is what I did.

  1. Reduced simultaneous motion flows and prioritized them
  2. Reduced the use of calling flows and especially those with “This flow is started with a text tag” and reduced text filtering (equals, contains)
  3. Reduced CPU intensive tasks of the app Device Capabilities (such as constantly showing a chronograph) or removed the old status field (deprecated). I’m waiting for the app to be migrated to the homey-api which uses less resources.
  4. Grouped all “repeat every” 5 min/30 min/60 min/1 day flows and let those flows be run sequentially instead of at the same time.
  5. Used helpers for time sensitive flows that have many conditional cards: A helper is set each time one of the conditions changes.
  6. I also used Advanced Triggers to ignore multiple motion events within a certain timeframe (debounce).

This brought down CPU load to a level of 20-30% and I have now great response times at any motion sensor. Hope this helps.